public async Task StopContinuousRecognitionAsync()
{
await speechRecognizer.StopContinuousRecognitionAsync().ConfigureAwait(false);
Logger.LogInformation($"Session stopped for {ConversationId}:{SpeakerType.ToString()}");
await OnCompletedSpeechAsync().ConfigureAwait(false);
}
IEnumerator Say(string speech, SpeakerType speaker = SpeakerType.mijin)
{
Debug.Log("[SpeechRenderrer::Say]" + speech);
//#if UNITY_ANDROID
// string uriSpeech = Application.persistentDataPath + "/tts.mp3";
//#else
// string uriSpeech = Application.dataPath + "/tts.mp3";
//#endif
// File.Delete(uriSpeech);
//ServicePointManager.ServerCertificateValidationCallback = Validator;
string url = "https://naveropenapi.apigw.ntruss.com/voice/v1/tts";
//HttpWebRequest request = (HttpWebRequest)WebRequest.Create(url);
//request.Headers.Add("X-NCP-APIGW-API-KEY-ID", "4lk8cmcq67");
//request.Headers.Add("X-NCP-APIGW-API-KEY", "Dnv1bksb2Trwh7DIbahih3QxFR9FOtAEdN1fPZz2");
//request.Method = "POST";
//byte[] byteDataParams = Encoding.UTF8.GetBytes("speaker=jinho&speed=0&text=" + speech);
//request.ContentType = "application/x-www-form-urlencoded";
//request.ContentLength = byteDataParams.Length;
//Stream st = request.GetRequestStream();
//st.Write(byteDataParams, 0, byteDataParams.Length);
//st.Close();
//HttpWebResponse response = (HttpWebResponse)request.GetResponse();
//string status = response.StatusCode.ToString();
////Console.WriteLine("status=" + status);
//using (Stream output = File.OpenWrite(uriSpeech))
//using (Stream input = response.GetResponseStream())
//{
// input.CopyTo(output);
//}
//WWW mp3Open = new WWW(uriSpeech);
//while (mp3Open.isDone)
//{
// yield return null;
//}
//byte[] mp3bytes = File.ReadAllBytes(uriSpeech);
//audioSource.clip = Utils.GetAudioClipFromMP3ByteArray(mp3bytes);
//audioSource.Play();
//StopCoroutine("Feedback");
//StartCoroutine("Feedback");
WWWForm form = new WWWForm();
Dictionary <string, string> headers = new Dictionary <string, string>();
headers.Add("X-NCP-APIGW-API-KEY-ID", "4lk8cmcq67");
headers.Add("X-NCP-APIGW-API-KEY", "Dnv1bksb2Trwh7DIbahih3QxFR9FOtAEdN1fPZz2");
//form.AddField("speaker", "jinho");
form.AddField("speaker", speaker.ToString());
form.AddField("speed", "0");
form.AddField("text", speech);
byte[] rawData = form.data;
using (WWW ttsRequest = new WWW(url, rawData, headers))
{
yield return(ttsRequest);
if (ttsRequest.error != null)
{
Debug.Log(ttsRequest.error);
}
audioSource.clip = GetAudioClipFromMP3ByteArray(ttsRequest.bytes);
audioSource.Play();
StopCoroutine("Feedback");
StartCoroutine("Feedback");
}
yield return(null);
}
public SpeechRecognition(IConfiguration config, ILogger <SpeechRecognition> logger)
{
if (config == null)
{
throw new ArgumentNullException(nameof(config));
}
if (logger == null)
{
throw new ArgumentNullException(nameof(logger));
}
this.Logger = logger;
this.Config = config;
var speechKey = config["Azure.Cognitive.Speech.Key"];
var speechRegion = config["Azure.Cognitive.Speech.Region"];
var speechLanguage = config["Azure.Cognitive.Speech.Language"];
var speechEndpointId = config["Azure.Cognitive.Speech.EndpointId"];
listSegment = new List <SegmentResult>();
speechConfig = SpeechConfig.FromSubscription(speechKey, speechRegion);
if (speechConfig != null)
{
speechConfig.SpeechRecognitionLanguage = speechLanguage;
speechConfig.OutputFormat = OutputFormat.Detailed;
if (!string.IsNullOrEmpty(speechEndpointId))
{
speechConfig.EndpointId = speechEndpointId;
}
PushStream = AudioInputStream.CreatePushStream();
audioConfig = AudioConfig.FromStreamInput(PushStream);
speechRecognizer = new SpeechRecognizer(speechConfig, audioConfig);
}
speechRecognizer.Recognizing += (s, e) =>
{
Logger.LogDebug($"Recognizing: Text={e.Result.Text}");
};
speechRecognizer.Recognized += (s, e) =>
{
if (e.Result.Reason == ResultReason.RecognizedSpeech)
{
// Building transcription.
var listBest = new List <NBest>();
foreach (var best in e.Result.Best())
{
listBest.Add(new NBest()
{
Confidence = best.Confidence,
Lexical = best.LexicalForm,
Itn = best.NormalizedForm,
MaskedItn = best.MaskedNormalizedForm,
Display = best.Text,
Sentiment = null,
Words = null
});
}
var segment = new SegmentResult()
{
ChannelNumber = null,
SpeakerId = (long)SpeakerType,
Offset = e.Result.OffsetInTicks,
Duration = e.Result.Duration.Ticks,
OffsetInSeconds = new TimeSpan(e.Result.OffsetInTicks).TotalSeconds,
DurationInSeconds = e.Result.Duration.TotalSeconds,
SpeechCompletedTimeStamp = DateTime.Now,
NBest = listBest.ToArray()
};
listSegment.Add(segment);
}
else if (e.Result.Reason == ResultReason.NoMatch)
{
Logger.LogDebug($"NoMatch: Speech could not be recognized.");
}
};
speechRecognizer.Canceled += (s, e) =>
{
Logger.LogInformation($"Canceled: Reason={e.Reason}");
if (e.Reason == CancellationReason.Error)
{
Logger.LogDebug($"Canceled: ErrorCode={e.ErrorCode}");
Logger.LogDebug($"Canceled: ErrorDetails={e.ErrorDetails}");
Logger.LogDebug($"Canceled: Did you update the subscription info?");
}
};
speechRecognizer.SessionStarted += (s, e) =>
{
Logger.LogInformation($"Session started for {ConversationId}:{SpeakerType.ToString()}");
};
speechRecognizer.SessionStopped += (s, e) =>
{
Logger.LogInformation($"Session stopped for {ConversationId}:{SpeakerType.ToString()}");
Logger.LogInformation("Stop recognition.");
};
}
public async Task OnCompletedSpeechAsync()
{
var culture = CultureInfo.InvariantCulture;
var timestamp = StartTime.ToString("s", culture);
var audioFileName = $"{ConversationId}/{ConversationId}-{culture.TextInfo.ToLower(SpeakerType.ToString())}-{timestamp}.wav";
var audioFileResult = new AudioFileResult()
{
AudioFileName = audioFileName,
AudioFileUrl = audioFileName,
SegmentResults = listSegment.ToArray()
};
var json = new RootObject()
{
AudioFileResults = new AudioFileResult[] {
audioFileResult
}
}.ToJson();
var localDirectory = Environment.GetEnvironmentVariable("LocalAppData");
var outFilePath = Path.Combine(localDirectory, $"{Guid.NewGuid()}.json");
try
{
File.WriteAllText(outFilePath, json);
if (File.Exists(outFilePath))
{
var blobName = $"{ConversationId}/{ConversationId}-{culture.TextInfo.ToLower(SpeakerType.ToString())}-{timestamp}.json";
await AzureStorageHelper.UploadTranscriptFileAsync(outFilePath, Config["Azure.Storage.ConnectionString"], Config["Azure.Storage.Container.Transcript"], blobName).ConfigureAwait(false);
File.Delete(outFilePath);
Logger.LogInformation($"Successfully uploaded transcript file for {ConversationId}:{SpeakerType.ToString()}.");
}
}
catch (IOException ex)
{
Logger.LogError(ex, $"Issue when uploading (or deleting) transcript file for {ConversationId}:{SpeakerType.ToString()}.");
};
}
public async Task ReceiveAsync(WebSocket socket, string conversationId, SpeakerType speakerType)
{
// PCM format, 16000 samples per second, 16 bits per sample, 1 channel (mono)
var outFormat = new WaveFormat(16000, 16, 1);
var localDirectory = Environment.GetEnvironmentVariable("LocalAppData");
var outFilePath = Path.Combine(localDirectory, $"{Guid.NewGuid()}.wav");
var startTime = DateTime.Now;
using (var outFileWriter = new WaveFileWriter(outFilePath, outFormat))
{
await speech.StartContinuousRecognitionAsync(conversationId, speakerType, startTime).ConfigureAwait(false);
var socketBuffer = new byte[Settings.ReceiveBufferSize];
if (socket != null)
{
var result = await socket.ReceiveAsync(new ArraySegment <byte>(socketBuffer), CancellationToken.None).ConfigureAwait(false);
while (!result.CloseStatus.HasValue)
{
outFileWriter.Write(socketBuffer, 0, result.Count);
if (result.Count > 0)
{
speech.PushStream.Write(socketBuffer, result.Count);
}
result = await socket.ReceiveAsync(new ArraySegment <byte>(socketBuffer), CancellationToken.None).ConfigureAwait(false);;
}
await speech.StopContinuousRecognitionAsync().ConfigureAwait(false);
await socket.CloseAsync(result.CloseStatus.Value, result.CloseStatusDescription, CancellationToken.None).ConfigureAwait(false);
}
outFileWriter.Close();
}
try
{
if (File.Exists(outFilePath))
{
var culture = CultureInfo.InvariantCulture;
var timestamp = startTime.ToString("s", culture);
var blobName = $"{conversationId}/{conversationId}-{culture.TextInfo.ToLower(speakerType.ToString())}-{timestamp}.wav";
await AzureStorageHelper.UploadAudioFileAsync(outFilePath, config["Azure.Storage.ConnectionString"], config["Azure.Storage.Container.Audio"], blobName).ConfigureAwait(false);
File.Delete(outFilePath);
logger.LogInformation($"Successfully uploaded audio file for {conversationId}:{speakerType.ToString()}.");
}
}
catch (IOException ex)
{
logger.LogError(ex, $"Issue when uploading (or deleting) audio file for {conversationId}:{speakerType.ToString()}.");
};
}
